Senior AI Engineer

Senior AI Engineer

AI Lab | Budapest, Hungary | Hybrid/Remote

You're not just writing code. You're building the intelligence layer of financial products that people trust with their money.

We're the AI Lab inside a Hungarian fintech that powers POS terminals and digital banking solutions. While others talk about AI, we ship it. Our systems process real transactions, handle real customer conversations, and make real decisions. Now we need engineers who can build both the brain and the body of these products.

What You'll Actually Do

This is a hybrid role for people who refuse to be boxed in. Some weeks you'll be knee-deep in RAG pipeline optimization. Others, you'll be building the React interface that makes that intelligence accessible to humans. We need T-shaped engineers who go deep in one area while staying dangerous across the full stack.

On the AI Side

• Design and build production RAG pipelines—chunking strategies, embedding models, vector databases (Pinecone, Weaviate, ChromaDB), and retrieval optimization
• Architect multi-agent systems using LangGraph and LlamaIndex for complex financial workflows
• Fine-tune open-source LLMs (Llama, Qwen, Mistral) for domain-specific tasks when API calls aren't enough
• Build AI microservices in Python/FastAPI that serve models with sub-second latency
• Implement evaluation frameworks (DSPy) to measure and improve model performance systematically

On the Product Side

• Build responsive frontends in React (Next.js) and TypeScript that make AI features intuitive
• Develop backend services in Node.js/NestJS or Python that orchestrate AI capabilities with business logic
• Own the full lifecycle: you build it, you deploy it, you monitor it, you improve it
• Work directly with product managers to translate business problems into technical solutions

Projects You'll Touch

Not hypotheticals—these are in production or shipping soon:

• Call Center Intelligence: Whisper-based transcription with LLM post-processing, task extraction, and Teams Bot integration
• Document Understanding: RAG pipelines with DSPy optimization for financial document Q&A
• Multi-Agent Workflows: Autonomous agents that handle complex, multi-step financial processes
• AI-Powered Internal Tools: Chat interfaces (Chainlit, custom React) that boost team productivity

Our Stack

• Cloud: Azure (AKS, Azure OpenAI Service), Docker, Terraform
• AI/ML: PyTorch, Hugging Face, LangChain, LlamaIndex, DSPy, Whisper
• Backend: Python (FastAPI), Node.js/TypeScript (NestJS)
• Frontend: React (Next.js), TypeScript
• Data: PostgreSQL, Vector DBs (Pinecone, Weaviate, ChromaDB)
• Observability: Prometheus, Grafana, Loki

What You Bring

Non-Negotiables

• 4+ years of production software engineering experience
• Strong Python skills (you can write clean, testable, production-ready code)
• Hands-on experience with at least one GenAI stack (LangChain, LlamaIndex, or similar)
• You've built and deployed web applications—frontend or backend, ideally both
• You can explain complex technical concepts to non-technical stakeholders

Bonus Points

• Experience with RAG systems in production (not just tutorials)
• You've fine-tuned LLMs for real use cases
• Background in fintech, banking, or high-stakes production systems
• MLOps experience (model monitoring, A/B testing, deployment pipelines)
• Contributions to open-source AI projects

This is NOT for You If...

We believe in radical honesty. This role isn't right for everyone:

• You prefer deep specialization. If you want to only do ML research or only write frontend code, we have separate roles for that.
• You need extensive guidance. We're a small team. You'll have support, but you'll also need to figure things out independently.
• You chase hype over substance. We adopt new tech when it solves real problems, not because it's trending on Twitter.
• You're not comfortable with ambiguity. GenAI is messy. Requirements change. We need people who thrive in uncertainty.
• You want a 9-to-5 where nothing is urgent. We move fast. Fintech has real deadlines and real consequences.

 Why Join Us

• Real Impact: Your code affects millions of transactions. This isn't a side project—it's core product.
• Ownership: Small team, big responsibility. No waiting months for approvals. Ship fast, learn faster.
• Modern Stack: We don't maintain legacy COBOL. You'll work with tools that didn't exist two years ago.
• No BS Culture: Flat hierarchy. Your ideas matter more than your title. We debate, decide, and move on.
• Budapest-Based Flexibility: Hybrid with actual flexibility. Come to the office when collaboration matters.

The Team

You'll join the Core AI Product Pod—a cross-functional team of about 10 people including GenAI Engineers, Technical Product Managers, and Platform Engineers. We operate with a "you build it, you run it" mentality and value craftspeople who take pride in their work.